A Parameter Identification Approach of a PEM Fuel Cell Stack Using Particle Swarm Optimization

The fear of fossil fuels depletion as well as the constantly increasing pollution rates motivated most of today’s engineers and researchers towards focusing on renewable energies and their applications. Fuel Cells are one of the green technologies that are being explored extensively around the world. The work of this paper was done on the 3kW ElectraGen™ fuel cell system under study for domestic use in the United Arab Emirates (UAE). Several experiments were conducted at different operating points and relatively high ambient temperatures. The experimental I/V characteristics of the system are matched by identifying 13 different modeling parameters using basic fitting. The obtained model is then further optimized using Particle Swarm Optimization (PSO). The resulting model is validated experimentally and was found to highly resemble the system’s I/V characteristics yielding less than 1.5 V H_∞ norm of the error.